SNOW HYDROLOGY IN A GENERAL-CIRCULATION MODEL

Citation
S. Marshall et al., SNOW HYDROLOGY IN A GENERAL-CIRCULATION MODEL, Journal of climate, 7(8), 1994, pp. 1251-1269
Citations number
52
Categorie Soggetti
Metereology & Atmospheric Sciences
Journal title
ISSN journal
08948755
Volume
7
Issue
8
Year of publication
1994
Pages
1251 - 1269
Database
ISI
SICI code
0894-8755(1994)7:8<1251:SHIAGM>2.0.ZU;2-B
Abstract
A snow hydrology has been implemented in an atmospheric general circul ation model (GCM). The snow hydrology consists of parameterizations of snowfall and snow cover fraction, a prognostic calculation of snow te mperature, and a model of the snow mass and hydrologic budgets. Previo usly, only snow albedo had been included by a specified snow line. A 3 -year GCM simulation with this now more complete surface hydrology is compared to a previous GCM control run with the specified snow line, a s well as with observations. In particular, the authors discuss compar isons of the atmospheric and surface hydrologic budgets and the surfac e energy budget for U.S. and Canadian areas. The new snow hydrology ch anges the annual cycle of the surface moisture and energy budgets in t he model. There is a noticeable shift in the runoff maximum from winte r in the control run to spring in the snow hydrology run. A substantia l amount of GCM winter precipitation is now stored in the seasonal sno wpack. Snow cover also acts as an important insulating layer between t he atmosphere and the ground. Wintertime soil temperatures are much hi gher in the snow hydrology experiment than in the control experiment. Seasonal snow cover is important for dampening large fluctuations in G CM continental skin temperature during the Northern Hemisphere winter. Snow depths and snow extent show good agreement with observations ove r North America. The geographic distribution of maximum depths is not as well simulated by the model due, in part, to the coarse resolution of the model. The patterns of runoff are qualitatively and quantitativ ely similar to observed patterns of streamflow averaged over the conti nental United States. The seasonal cycles of precipitation and evapora tion are also reasonably well simulated by the model, although their m agnitudes are larger than is observed. This is due, in part, to a cold bias in this model, which results in a dry model atmosphere and enhan ces the hydrologic cycle everywhere.